Selective ratiometric detection of H₂O₂ in water and in living cells with boronobenzo[b]quinolizinium derivatives

Abstract

Boronobenzo[b]quinolizinium derivatives exhibit several favorable properties for the fluorimetric detection of hydrogen peroxide, namely quantitative transformation to a product whose emission maximum is well separated from the one of the substrate, water solubility, and the ability to operate in living cells.

---

Reactive oxygen species (ROS), such as e.g. hydrogen peroxide, singlet oxygen, the hydroxyl radical, the superoxide anion, peroxynitrite, or ozone,¹ have fundamental functions in health and disease related processes.² Specifically, increased concentrations of H₂O₂ may cause oxidative damage of cellular proteins, ageing and diseases such as cancer or diabetes.³ Therefore, the detection of H₂O₂ and other ROS is a challenging and important task in biological and chemical research. Along these lines, fluorescent probes have been established as sensitive and selective tools for the detection of H₂O₂.^4a Fluorescent light-up probes offer one approach to respond to H₂O₂ by an emission increase.⁵ These intensity-based probes may have practical advantages; however, they suffer from the fact that the emission intensity is also effectively influenced by other internal and external factors.^4b A complementary approach towards fluorimetric detection of ROS is based on ratiometric probes operating at two different detection wavelengths that enable an accurate internal calibration.⁶ In this context aryl boronates were identified as probe molecules whose selective reaction with H₂O₂ may be employed for the development of ratiometric fluorescent probes.⁷ In general the transformation of the boronic acid (ester) group into an hydroxy functionality results in a significant change of the absorption and emission properties of the fluorophore thus enabling a ratiometric analysis. Based on this principle, several ratiometric H₂O₂-sensitive probes were developed.⁸ Nevertheless, the development of novel ratiometric fluorescent chemosensors for the detection of H₂O₂ is still an important goal, especially those with sufficient water solubility. The latter is an essential requirement for studies in biological media. In this context we have established the benzo[b]quinolizinium as a versatile, water-soluble fluorophore for fluorimetric detections.⁹ In addition, we have discovered that the hydroxy-substituted benzo[b]quinolizinium derivative 2b has fluorosolvatochromic properties with a large redshift of the emission band in water (600 nm).¹⁰ At the same time, the 9-boronobenzo[b]quinolizinium (1a) exhibits the same emission bands as the parent compound (417 nm).¹¹ Considering this large difference of emission color of borono- and hydroxy-substituted derivatives we proposed that the formation of the latter by the reaction of H₂O₂ with boronobenzo[b]quinolizinium derivatives may be monitored by a drastic shift of the emission wavelength. Meanwhile, it was shown that resembling quinolinium-type boronic acid esters may indeed be used for that purpose.^8a Herein, we will demonstrate that the boronic acids and ester derivatives 1a–d may be generally employed as ratiometric fluorescence probes for the detection of H₂O₂ in aqueous solution and in living cells (Scheme 1).

Scheme 1 Reaction of boronobenzo[b]quinolizinium derivatives 1a–d with H₂O₂.

The benzo[b]quinolizinium derivatives 1a,b and 2a,b were synthesized by the cyclodehydration method,¹² and the two acid derivatives 1a,b were transformed to the corresponding pinacol-boronates 1c,d (cf. ESI†). Known compounds (1a, 2a,b) were identified by comparison with literature data.^10,11,12b All new compounds (1b–d) were characterized by ¹H-NMR and ¹³C-NMR spectroscopy, mass spectrometry and elemental analysis (cf. ESI†). The absorption spectra of the derivatives 1a–d in phosphate buffer solution reveal the characteristic long-wavelength absorption band of the benzo[b]quinolizinium chromophore¹³ with a maximum at 380 nm (1a,c) and 377 nm (1b,d) (cf. ESI†). The hydroxy-substituted derivatives have absorption maxima at 387 nm (2a) and 379 nm (2b),¹⁰ along with a weak, very broad redshifted charge transfer (CT) band at ca. 450 nm.¹⁰ The emission properties of the boronic acid derivatives 1a–d resemble the ones of the parent benzo[b]quinolizinium (λ_fl = 410–415 nm; Φ_fl = 0.03–012). In contrast, the fluorescence maxima of the hydroxybenzo[b]quinolizinium derivatives are significantly redshifted with maxima at 527 nm (2a) and 600 nm (2b).¹⁰

The reaction of the derivatives 1a–d with H₂O₂ in buffer solution produced the corresponding hydroxybenzo[b]quinolizinium derivatives 2a and 2b (Scheme 1). Most notably, in the case of 1a–d the ¹H-NMR-spectroscopic analysis of the reaction revealed a quantitative transformation with no detectable side products (cf. ESI†). The reaction of compounds 1a–d with H₂O₂ in buffer solution was also monitored by emission spectroscopy (Fig. 1). The progress of the reaction is clearly indicated by a continuous decrease of the intensity of the emission maximum of compounds 1a–1d (415 or 417 nm) that is accompanied by the simultaneous increase of a new redshifted emission band at 527 nm and 600 nm, respectively. The latter emission bands are identical to the ones obtained with authentic samples of the hydroxy-substituted derivatives 2a,b. The ratios of the intensities of the emission maxima increase by a factor of 17 (1a) and 0.7 (1b) at 140 min after addition of H₂O₂, as quantified by the term I_red/I_blue. Control experiments showed that the formation of 2a and 2b does not take place in buffer solution in the absence of H₂O₂. The results indicate full conversion of the substrates 1a,b to the products 2a,b within 140 min (1a) and 160 min (1b). The limit of detection was determined to be 3.0 μM for 1a and 5.9 μM for 1c (cf. ESI†), which is comparable to that of established probes.^5b,8a,c The reaction rate was analyzed considering pseudo first-order kinetics according to an established protocol.^8c By this method, rate constants of k_obs = 4.3 × 10⁻⁴ s⁻¹ (1a) and k_obs = 2.9 × 10⁻⁴ s⁻¹ (1b) were obtained (cf. ESI†). To assess the selectivity of the oxidation reaction of compounds 1a–d they were also treated with other oxidants, namely O₂˙⁻, ¹O₂, ClO⁻, NO˙, peroxides ROO˙, HO˙, and ONO₂⁻. The reaction was monitored by fluorescence spectroscopy, and the ratiometric analysis, I_red/I_blue, was used to determine the progress of the reaction (Fig. 2). Notably, the compounds 1a and 1c are efficiently oxidized by H₂O₂, whereas the other oxidants only induce the formation of the oxidation product 2a to marginal or moderate (O₂˙⁻, HO˙, ONO₂⁻) extent under resembling reaction conditions. A similar trend was observed for the derivatives 1b and 1d, however, the selectivity is less pronounced, and the ratio of I_red/I_blue is smaller because of the lower emission intensity of the reaction product 2b.

Fig. 1 Fluorimetric monitoring of the reaction of H₂O₂ (1.0 mM) with 1a (A, 10 μM, λ_ex = 366 nm), 1b (B, 10 μM, λ_ex = 361 nm), 1c (C, 10 μM, λ_ex = 366 nm), or 1d (D, 10 μM, λ_ex = 366 nm) in phosphate buffer (0.1 M, pH = 7.4). Arrows indicate the development of emission bands with time. Inset: plot of the fluorescence maxima of 1a–d (●) and 2a,b (○) versus time.

Fig. 2 Ratiometric fluorimetric analysis of the reaction of 1a (A, black), 1c (A, grey), 1b (B, black), and 1d (B, grey) (c = 10 μM) with selected ROS (c = 1.6 × 10⁻⁴ M, resp.) in phosphate buffer (c = 0.1 M, pH = 7.4); t = 60 min; 1a, 1c: λ_ex = 366 nm, 1b, 1d: λ_ex = 361 nm; ROO˙ = C₄H₉N₂OO˙.

The ability of the derivatives 1c and 1d to detect varying levels of H₂O₂ in living cells was assessed. For this purpose, we used HeLa cell lines, i.e. an adenocarcinoma that grows as monolayer, and Jurkat cells, i.e. leukemia cell lines that grow in suspension. In preliminary experiments the cellular uptake by compound 1c was evaluated by flow cytometry taking advantage of the intrinsic fluorescence of the compound (cf. ESI†). The cellular uptake was rapid and reached a plateau after 15–20 min of incubation in Jurkat cells and after 1 h in HeLa cells suggesting an efficient delivery of the compound inside the cells. After 1 h of incubation 1c and 1d emit at 450 nm (FL9 channel), while no emission was detectable at 540 nm (FL10 channel) (cf. ESI†). To evaluate the ability of derivatives 1c and 1d to respond to H₂O₂ in a living cell, HeLa cells were treated with H₂O₂ (100 μM) after the incubation with 1c and 1d, and the relative fluorescence intensities of the cells were analyzed by flow cytometry (Fig. 3). After 10 min the fluorescence intensity of 1c at 450 nm decreased, and a signal at 540 nm developed (Fig. 3A and B). This development of emission signals may be followed also by a ratiometric analysis (cf. ESI†). In the case of 1d, however, only the signal at 450 nm disappeared, but no redshifted emission was observed.

Fig. 3 Flow cytometric analysis of HeLa cells incubated with 1c (A and B) and 1d (C and D) for 1 h (dotted lines) treated with 100 μM H₂O₂ and then re-analyzed after 10 min of incubation (continuous lines). The gray shadows represent untreated cells. The data represent at least 10 000 cells for each analysis.

Furthermore, we evaluated if 1c could be useful to detect physiological generation of H₂O₂ under pharmacological induction. Thus, both HeLa and Jurkat cells were treated with staurosporine, that is known to induce apoptosis through mitochondrial membrane depolarization. The latter has been associated with mitochondrial production of reactive oxygen species (ROS), in particular H₂O₂.¹⁴ HeLa and Jurkat cells were treated with the drug for 12 h and afterwards incubated with 1c for 30 min. Subsequent analysis by flow cytometry showed a significant increase of the population of cells that emit at 540 nm in both cell lines (Fig. 4), thus demonstrating the capability of compound 1c to indicate pharmacologically stimulated formation of intracellular H₂O₂.

Fig. 4 Flow cytometric analysis of HeLa (A) and Jurkat cells (B) treated for 24 h with staurosporine (1 μM) and subsequently incubated with 1c for 20 min.

In summary we have demonstrated that boronobenzo[b]quinolizinium derivatives exhibit several favorable properties for fluorimetric detection of hydrogen peroxide, namely water solubility, quantitative transformation to a product whose emission maximum is well separated from the one of the substrate thus allowing ratiometric analysis, and the general ability to operate in living cells. With respect to a relationship between function and structure, it appears that benzo[b]quinolizinium fluorophores, in which the ROS-sensitive borono-functionality as well as the eventually formed hydroxy functionality are linearly conjugated with the pyridinium unit, provide the more effective structural platform for fluorimetric detection as compared to the cross-conjugated system. Specifically, the 9-substituted derivatives 1a and 1c exhibit a more pronounced light-up effect and a higher selectivity towards H₂O₂ than the 8-substituted ones. It should be noted that these probes may be easily varied by the attachment of other functionalities R¹ that are transformed to the strongly solvatochromic hydroxy-substituted derivative 2a.

Generous support by the Deutsche Forschungsgemeinschaft is gratefully acknowledged.

Notes and references

1. M. P. Fink, Curr. Opin. Crit. Care, 2002, 8, 6–11.
2. (a) P. Niethammer, C. Grabher, A. T. Look and T. J. Mitchison, Nature, 2009, 459, 996–999; (b) G. Groeger, C. Quiney and T. G. Cotter, Antioxid. Redox Signaling, 2009, 11, 2655–2671; (c) E. A. Veal, A. M. Day and B. A. Morgan, Mol. Cell, 2007, 26, 1–14; (d) J. Li, M. Stouffs, L. Serrander, B. Banfi, E. Bettiol, Y. Charnay, K. Steger, K. H. Krause and M. E. Jaconi, Mol. Biol. Cell, 2006, 17, 3978–3988; (e) M. Ushio-Fukai, Cardiovasc. Res., 2006, 71, 226–235; (f) E. R. Stadtman, Free Radical Res., 2006, 40, 1250–1258.
3. D. Hernández-García, C. D. Wood, S. Castro-Obregón and L. Covarrubias, Free Radical Biol. Med., 2010, 49, 130–143.
4. (a) C. C. Winterbourn, Biochim. Biophys. Acta, 2014, 1840, 730–738; (b) Y. H. A. Liu and X. B. Liao, Curr. Org. Chem., 2013, 17, 654–669.
5. See e.g. (a) M. Kaur, D. S. Yang, K. Choi, M. J. Cho and D. H. Choi, Dyes Pigm., 2014, 100, 118–126; (b) K. B. Daniel, A. Agrawal, M. Manchester and S. M. Cohen, ChemBioChem, 2013, 14, 593–598; (c) L. Yuan, W. Lin, Y. Xie, B. Chen and S. Zhu, J. Am. Chem. Soc., 2012, 134, 1305–1315; (d) Y. Ahn, K. E. Fairfull-Smith, B. J. Morrow, V. Lussini, B. Kim, M. V. Bondar, S. E. Bottle and K. D. Belfield, J. Am. Chem. Soc., 2012, 134, 4721–4730; (e) L. Yuan, W. Lin, S. Zhao, W. Gao, B. Chen, L. He and S. Zhu, J. Am. Chem. Soc., 2012, 134, 13510–13523; (f) N. Karton-Lifshin, E. Segal, L. Omer, M. Portnoy, R. Satchi-Fainaro and D. Shabat, J. Am. Chem. Soc., 2011, 133, 10960–10965.
6. (a) C.-Y. Chen and C.-T. Chen, Chem. – Eur. J., 2013, 19, 16050–16057; (b) A. R. Lippert, G. C. V. De Bittner and C. J. Chang, Acc. Chem. Res., 2011, 44, 793–804; (c) A. E. Albers, V. S. Okreglak and C. J. Chang, J. Am. Chem. Soc., 2006, 128, 9640–9641.
7. (a) Y. H. A. Liu and X. B. Liao, Curr. Org. Chem., 2013, 17, 654–669; (b) Z. Q. Guo, I. Shin and J. Yoon, Chem. Commun., 2012, 48, 5956–5967; (c) For mechanism, see: H. G. Kuivila and A. G. Armour, J. Am. Chem. Soc., 1957, 79, 5659.
8. (a) S. W. Lee, H.-W. Rhee, Y.-T. Chang and J.-I. Hong, Chem. – Eur. J., 2013, 19, 14791–14794; (b) G. P. Li, D. J. Zhu, Q. Liu, L. Xue and H. Jiang, Org. Lett., 2013, 15, 924–927; (c) G. Masanta, C. H. Heo, C. S. Lim, S. K. Bae, B. R. Cho and H. M. Kim, Chem. Commun., 2012, 48, 3518–3520; (d) D. Srikun, E. W. Miller, D. W. Dornaille and C. J. Chang, J. Am. Chem. Soc., 2008, 130, 4596–4597; (e) A. E. Albers, V. S. Okreglak and C. J. Chang, J. Am. Chem. Soc., 2006, 128, 9640–9641.
9. (a) R. Bortolozzi, H. Ihmels, L. Thomas, M. Tian and G. Viola, Chem. – Eur. J., 2013, 19, 8736–8741; (b) M. Tian, H. Ihmels and S. T. Ye, Org. Biomol. Chem., 2012, 10, 3010–3018; (c) M. Tian, H. Ihmels and K. Benner, Chem. Commun., 2010, 46, 5719–5721; (d) M. Tian and H. Ihmels, Chem. Commun., 2009, 3175–3177; (e) A. Bergen, A. Granzhan and H. Ihmels, Photochem. Photobiol. Sci., 2008, 7, 405–407.
10. H. Ihmels and K. Schäfer, Photochem. Photobiol. Sci., 2009, 8, 309–311.
11. M. Tian and H. Ihmels, Synthesis, 2009, 4226–4234.
12. (a) C. K. Bradsher, Chem. Rev., 1946, 46, 447–499; (b) C. K. Bradsher and J. H. Jones, J. Am. Chem. Soc., 1957, 79, 6033–6034.
13. S. U. D. Saraf, Heterocycles, 1981, 16, 987–1007.
14. (a) H. Nohl, L. Gille and K. Staniek, Biochem. Pharmacol., 2005, 69, 719–723; (b) J. Cai and D. P. Jones, J. Biol. Chem., 1998, 273, 11401–11404.

---

Footnotes

† Electronic supplementary information (ESI) available: Experimental procedures, full characterization and NMR spectra, spectroscopic studies. See DOI: 10.1039/c4cc02283a

‡ Author names in alphabetical order that does not reflect the specific contribution of each author.

---